Keeper AI Test and Security Posture: What the Linked Site Shows

The linked page does not establish that cybersecurity capability. The visible keeperaitest.com page is framed as a dating-standards or “AI Matchmaker” calculator, while its prominent calls to action navigate to crushon.ai. That page position is not the same as improving an organization’s cybersecurity posture. It also does not prove that a calculator is fully runnable on the exact page, that inputs stay local, or that any enterprise security feature exists.

An illustration does not establish threat-detection, encryption or access-control capabilities, or verify that a calculator works.

The linked domain is Keeper AI Test. The visible copy says the page helps estimate how many people may fit selected relationship preferences, using demographic-style criteria such as age, height, income, religion, or ethnicity. That identity should not be merged with a password manager, SOC tool, SIEM, endpoint product, or security-posture-management platform simply because the name includes “Keeper” or “AI.”

Why this is not “security posture”

Security posture usually means the state of an organization’s assets, controls, detection, logging, vulnerabilities, response readiness, and governance. A dating-preference page does not demonstrate those functions. CISA’s Secure by Design guidance says customers need ways to evaluate a software maker’s security progress and emphasizes transparency and accountability: CISA Secure by Design. For Keeper AI Test, the missing evidence would be items such as a control list, architecture diagram, audit report, vulnerability disclosure policy, access logs, data retention schedule, and incident-response commitments.

Population statistics also do not create a cybersecurity result. Even if a page cites the U.S. Census Bureau or CDC, a data-source label does not prove that the model combines variables correctly, that variables are independent, that the sample is current, or that the result predicts an individual user’s chance of finding a partner. A percentage of a demographic pool is not a breach-risk score, authentication control, or monitoring signal.

How to verify “no personal data stored” without overclaiming

The page’s privacy-style wording is useful only if the scope is clear. “No personal information is stored” might mean no profile is saved by the calculator, but it might not cover server access logs, IP addresses, cookies, analytics events, referral URLs, crash logs, CDN logs, or data sent after clicking a third-party CTA. NIST describes the Privacy Framework as a tool for identifying and managing privacy risk while protecting individuals’ privacy: NIST Privacy Framework. That means the claim should be tested by scope, not repeated as a blanket guarantee.

1. Use fictional inputs only. Do not enter a real name, precise income, real address, employer, or contact details. Use broad synthetic values and record the page state.
2. Check storage in the browser. In a clean browser profile, inspect cookies, local storage, and session storage before and after interaction. This can show client-side persistence, but it cannot reveal server logs.
3. Check network destinations. With developer tools open, observe whether form changes or button clicks send values to keeperaitest.com, crushon.ai, analytics domains, ad domains, or other scripts.
4. Follow the CTA boundary. If a button leaves the domain, treat the destination as a separate service with separate terms, cookies, account handling, and retention.
5. Ask for written scope. A useful answer states whether IP logs, user-agent strings, click events, referral parameters, and abuse-prevention logs are stored, and for how long.

A practical conclusion is narrower and more accurate: keeperaitest.com may be a relationship-standards information page with privacy claims to verify. It does not document a cybersecurity-posture product, and its demographic percentages should not be interpreted as individual matchmaking certainty or as security evidence.